
Meeting in Automation Lab - Chem. Engg. 29/07/2010

present: Prof. Kannan Moudgalya[KM], Prof. Pradeep Sarin[PS], Aditi Rathi, Abhishek Pawar, Mehul. 
Prashant and Saurabh (EE) joined a little late.
C. P. Varun (B.Tech, Engg. Physics - 4th year) was introduced as new member to the group. He has 
substantial experience in embedded hardware systems, have worked on developing a micro-controller based
Braille E-Book reader last year.

minuted by Pradeep Sarin

--------------------------------------

Focus of the meeting was to discuss the push from MHRD-Mr. Sinha to move in the direction of 
HawkBoard development.

1. Aditi gave a status update on what's been happening on the systems programming front. Most of this
   already up on Wiki page (status updates) section - so I'm not repeating it here.

2. KM related his discussions with Mr. Sinha of the previous day regarding HawkBoard initiative:
   - MHRD feels confident being able to procure HawkBoard for $10 (processor component $5)
   - display for $10. Battery power $5-$10 leaving the rest $5-$10 for systems integration cost.

3. Group discussed at length how the $10 Hawkboard cost was arrived at. he cost through retail channels 
   listed on http://www.hawkboard.org is ~$100
   example: http://us.element-14.com/innovate-software-solutions/issplhawk/hawk-board-omap-l138/dp/65R5489?
   
   Our conclusion was that the assumed $10 cost must be for $5 processor subsidized from TI, and an
   absolutely stripped down bare-bones version of the open-hardware HawkBoard developed in-house.

4. Display cost is dependent upon the resolution and use/non-use of a touch-screen. There was an extended
   debate on the usability of touch-screen v/s qwerty keyboard. PS has posted some detailed studies data
   on the Wiki page. The key issues involved are: power consumption, ruggedness, ergonomics. The monetary 
   cost may not be much of an issue, but power consumption and hardware complexity may turn out to be
   the key deciding factors.

5. Action Item decided upon:---(directive from MHRD) need to strip down OS to bare essentials and test
   HawkBoard performance on delivering the content the LCAD is designed for. This must involve:
   a) software benchmarking (PS has put up links to benchmarking protocols/sw on the wiki)
   b) power consumption under load.
   KM proposed that we procure about 10 of the HawkBoard units and offer them on an IIT-wide competition
   for systems programming to achieve these objectives.

6. Procurement of HawkBoards:
   a) Current vendor has no stock - will be able to deliver 10 only by August 30 at the earliest.
   b) We decided to proceed with in-house fabrication of the boards using the open-hardware schematics.
      Rationale is that we will have to do this anyway to strip down the hardware to bare minimum 
      components. 
      Rajesh Kushalkar can liasion with TI hardware systems team to validate our design at the schematic
      and PCB level.
      4-6 layer PCB's can be fabricated using local Bombay based vendors (Abhishek,Varun to co-ordinate)
      PCB stuffing with SMD components can be done in EE department or outsourced if needed.
      Abhishek to look into procuring the component kits for stuffing 10x PCB's (DigiKey and/or local 
      Lamington Rd vendors).

------------------------------------

As a fun pre-meeting exercise, we did a coarse measurement of the power consumed
 by the HawkBoard by connecting an ammeter in series with the DC power adapter
 connection. At idle, running X-windows driving the HDMI video output to a 
 regular desktop LCD monitor (640x480 res), it consumes ~ 0.3-0.4 amps at 5V,
which implies ~ 2 W. 
Note that the monitor isbeing driven seperately with it's own AC power 
connection, so this power does not budget for a monolithic device with a 
touchscreen/LCD powered by the same source as the HawkBoard itself.

  


